See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/313940662

Comprehensive assessment of virtual synchronous machine based voltage

source converter controllers

Article in IET Generation, Transmission & Distribution · February 2017

DOI: 10.1049/iet-gtd.2016.1423

CITATIONS READS

94 1,144

2 authors, including:

Hasan Alrajhi
Umm Al-Qura University
23 PUBLICATIONS 281 CITATIONS

SEE PROFILE

All content following this page was uploaded by Hasan Alrajhi on 10 December 2019.

The user has requested enhancement of the downloaded file.

IET Generation, Transmission & Distribution

Research Article

Comprehensive assessment of virtual ISSN 1751-8687

Received on 7th September 2016
Revised 28th December 2016
synchronous machine based voltage source Accepted on 15th February 2017
E-First on 24th April 2017
converter controllers doi: 10.1049/iet-gtd.2016.1423
www.ietdl.org

Hasan Alrajhi Alsiraji1,2 , Ramadan El-Shatshat1

1Department of Electrical and Computer Engineering, University of Waterloo, 200 University Ave W, Waterloo, ON, Canada
2Department of Electrical Engineering, Umm al-Qura University, Al-Taif Road, Mecca 24382, Saudi Arabia
E-mail: halsiraj@uwaterloo.ca

Abstract: The substantial potential for the integration of renewable energy into power systems using power electronics
converters might result in stability issues because of a lack of inertia. For this reason, this study introduces the concept of a
virtual synchronous machine (VSM) control algorithm that emulates the properties of traditional synchronous machines. The
literature includes references to several differently structured control algorithms. However, synchronous machine inertia and
damping characteristics must be mimicked, which makes the cost and simplicity of implementation important from an economic
perspective. This study presents a comprehensive comparison of VSM control algorithms. The most significant factor
investigated in the work presented in this study is the viability of VSM algorithms during the kind of abnormal operation that
might raise instability issues with respect to practical discrete time operation. The test system used in this study, which was
simulated in a PSCAD/EMTDC environment, consisted of simulated voltage source converters based on a fully detailed
switching model with two AC voltage levels. The results indicate a significant outcome that can facilitate a determination of the
most effective VSM control algorithm.

1 Introduction behaviour of an SM. In [6], the implementation of virtual inertia in

To meet increased electrical power demands and address
environmental concerns, renewable energy sources (RESs) are
being integrated into traditional AC systems because of their
potential to diminish the need for energy derived from fossil fuels
and decrease greenhouse gas emissions. Unfortunately, RESs
cannot fit seamlessly into conventional power grids that are
designed without consideration of distributed generation. The
random power variations inherent in weather-based generation
create varying degrees of uncertainty that affect numerous system
variables, including bus voltages and the system frequency [1].
High levels of RES penetration may also lead to stability problems
and affect system dynamics [2, 3]. Due to their numerous benefits
and the superior performance they offer, voltage source converters
(VSCs) have become a practical way to interface renewable
sources characterised by either variable frequencies or direct
current, such as wind, photovoltaic, and fuel cells. However, unlike
synchronous machines (SMs), the VSCs that interface with RESs
lack the inertia needed to support and participate in the frequency
and voltage control of an AC system and are hence unable to
contribute to the improvement of system stability.
This issue can be addressed through the application of the
concept of virtual SM (VSM) controllers, which combine VSC and
SM characteristics. The VSM control algorithm of a power
electronics converter is a control feature that can be added to a
converter controller to enable it to behave as an SM [4]. The
algorithm allows the incorporation of virtual inertia and damping
behaviour into the loop controllers of an interfacing converter [5,
4]. The VSC thus has no physical mass of inertia as in the case of
practical SMs, i.e. it does not have limited inertia or constant
damping values. Instead, only the converter power response
mimics the inertia response of a real SM.
VSM control algorithms, which can be implemented through a
slight modification to the VSC controller, are categorised as either
high-order or low-order VSM control algorithms as shown in
Fig. 1. Although both types of VSM algorithms provide active and
reactive power control, each has a specific control frame, as
discussed in detail in the next section. A number of VSM control
algorithms have been developed to enable a VSC to mimic the
the converter controller loops with a storage unit was introduced as
a virtual synchronous generator (VSG). Chen et al. [2] presented a
different VSM concept, which they represented as a hysteresis
current-controlled SM model for controlling a three-phase inverter;
they named it VISMA. The control of a VSG that emulates SM
transient as well as steady-state characteristics was further
investigated in [5, 7], while a VISMA algorithm was adopted by
Chen et al. [8]. As an alternative, synchronverters were employed
in the work reported in [9, 10], and a theoretical review of these
control algorithms was provided in [4, 11]. These types of VSMs
can be classified in terms of either their output references [11] or
their topologies [4]. The majority of the studies related to VSM
control algorithms that are described in the literature have been
focused on their development and implementation. In [12], the
dynamic characteristics of VSGs were compared with droop
control under normal operating conditions. Tande [13] examined
the inertial dynamics of a VSG used with doubly fed induction
generator (DFIG) based wind turbines by including a short circuit
in the AC-side system, which is remote from the converter. A VSG
for grid-connected DFIG-based wind turbines was studied with
respect to different damping coefficients and varied short circuit
ratios [14]. In other work [10], the performance of a hybrid multi-
terminal direct current was evaluated under single-fault conditions
with a VSM-based synchronverter.
The choice of an appropriate VSM control algorithm (low- or
high-order) to be implemented is critical, far from clear-cut, and
still requires further investigation. Only limited comparative
studies of VSM algorithms are available [15, 16]. Arco and Suul
[15] studied the equivalence of a VSM for conventional frequency
droop control and compared it with VSC droop control in order to
demonstrate similar behaviour during load step changes. Another
study focused on the difference between the implementation of two
types of high-order SM models with respect to implementation
simplicity and the behaviour of the converter with balanced AC
voltages [16]. All of these previous studies involved the
investigation and comparison of VSM controllers based on an ideal
voltage source. An important element of determining the most
appropriate algorithm is therefore to examine and compare the
dynamic characteristics and behaviour of VSM control algorithms

IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769 1762
© The Institution of Engineering and Technology 2017
Fig. 1 Classification of VSM based on model's order
and their performance during the abnormal conditions that
commonly occur in distribution systems, such as short circuits and
unbalanced AC voltages.
The majority of the previous works in the literature were
focused on the implementation concept of VSM, but, they did not
carried out the serious issues as an unbalance AC voltage and short
circuit situation. This paper presents a comprehensive examination
and assessment of VSM control algorithms in order to establish
suitable choices of VSM control algorithms for implementation
and use in a variety of applications. Therefore, this paper revealed
the serious issue with implementing the high-order model in the
practical real system; which might lead to unstable operation
especially in AC microgrid application. The study presented in this
paper also entailed an investigation of the equivalence of a low-
order VSM control algorithm with a high-order VSM control
algorithm under normal and abnormal conditions, such as
unbalanced AC voltages and short circuits. A comparison of the
dynamic and performance behaviour of VSM control in grid-
connected mode for different operating scenarios is provided.
However, since conventional frequency droop control provides
behaviour equivalent to that of a VSM [15], consideration of both
algorithms in this study was unnecessary because conventional
frequency droop control lacks an inertia coefficient in the control
loops, which means zero inertia [15]. The system discussed in this
paper is based on a detailed switching VSC model, which was
simulated in a PSCAD/EMTDC environment.
The rest of the paper is organised as follows. Section 2 provides
a brief overview of VSM control algorithms and their
categorisation, Section 3 presents the simulation results and
analysis, and the final section offers conclusions.
2 VSM review
The nature of RES is not deterministic, so the power will fluctuate;
therefore, the voltage level and the system's frequency can be
affected [17]. Solving this issue can be done via adding an extra
inertia to the system [4]. Therefore, the operation of several SMs is
a hint of combining power electronics with the behaviour of the
SMs to be emulated [18].
There are enormous applications for VSM control concept.
Using the second-order VSM model for smart grid application was
IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769
© The Institution of Engineering and Technology 2017
studied to show the seamless transition between grid-connected
mode and islanded mode [19]. Moreover, applying the VSM based
on high-order model was investigated for parallel inverter in AC
microgrid [20]; however, this model does not provide the converter
either overloading or overcurrent protection. Grid-connected
DFIG-based wind turbines were studied with high-order VSM
including currents limitation [14], but the model mainly depends
on measuring the instantaneous voltages. Therefore, unbalance
voltages and short circuit conditions affect the instantaneous
measured voltages for the high-order VSM model, which are not
yet investigated in the literature. These conditions will affect the
system stability. The power distribution system with high
penetrations of converters was studied with including the VSM
control concept to improve system performance [21]. Nevertheless,
the VSM control concept can be implemented for single-phase
bidirectional battery charger for providing vehicle-to-grid services
[22]. High-voltage DC application connected to weak AC system
using the VSM control concept offers frequency support to the
weak AC system with low inertia [23]. Moreover, in AC microgrid
application, the transition from islanded mode to grid-connected
mode presents large transients in voltages and currents that reflects
high transient in power might hit the stability of the microgrid.
Using the concept of VSM control provides smooth transition of
the AC microgrid operation [24].
Due to the existence of several VSM control algorithms for
VSC, they can be classified into two categories which are high-
order and low-order VSM. Also, the VSM models can be divided
based on the requirement of an extra component such as energy
storage; therefore, Fig. 1 summarises the differences of VSM
models at different stages, and they were found in the literature.
Therefore, the choice of the VSM models to be assessed and
studied is based on the fact that there is no need of an extra
component as shown in Fig. 1 in the light blue box. However, the
VSM based on phasor SM is not considered due to a serious
problem when it is implemented in d–q frame that is the model
become unsteady state in case of unbalanced load [8]. Moreover,
the droop control does not have yet virtual inertia unless it includes
a low-pass filter (LPF) to filter out the active power. It is clear that
without LPF the droop control has zero inertia [15].
1763

Fig. 2 Simplified SM model
2.1 High-order VSM model [4]
Mimicking the behaviour of SM inertia on VSC control algorithm
can be achieved via introducing virtual inertia and damping
characteristics into VSC control using the mathematical model of
SM [4–7]. Subsequently, the idea of implementing a virtual inertia
to the converter controller loops with storage unit is introduced as
VSG [5], which requires an extra storage unit. Also, Chen et al. [2]
presented another concept of VSM (VISMA) which is represented
as hysteresis current controlled of the high order of traditional SM
model to control the three-phase inverter. Moreover, in [16], the
authors presented and investigated the behaviours of two high-
order models of VSM based on current-to-voltage and voltage-to-
current of SM model, so they used different control techniques
which are a pulse-width modulation (PWM) controller and a
hysteresis controller, respectively.
The high-order model of the SM is used to calculate the
reference values. It is worthwhile to notice that the mathematical
model of SM consists of mechanical and electrical parts. The
mechanical part is responsible for generating the angular speed and
a positive phase voltage angle which is known as swing equation.
The electrical part is the stator winding equations which are
energised by the phase angle to generate either the current or
voltage references to the controller. The interaction between the
mechanical and electrical parts is simplified as shown in Fig. 2, and
(1)–(5) are based on a SM simplified model which is retrieved
from [2]
dia
Ea − va = Rsa ⋅ I a + Lsa ⋅ (1)
dt
dib
Eb − vb = Rsb ⋅ I b + Lsb ⋅ (2)
dt
dic
Ec − vc = Rsc ⋅ I c + Lsc ⋅ (3)
dt
1 dw
= Pmech − Pele − Pd (4)
J dt
θ= ∫ w ⋅ dt (5)
In (1)–(3) where Ea, Eb, and Ec are the voltages at point of
common coupling (PCC) which are measured values; va, vb, and vc
are the induced back electromotive forces; the symbols Rs and Ls
represent the stator resistances and the stator inductances,
respectively. In (4), the machine's inertia is denoted by J; the
mechanical input power is denoted by Pmech, and the electrical
power is represented Pele. The damping power is denoted by Pd. In
1764
(5), the angular speed is denoted by w and a positive phase voltage
angle is represented by θ.
The high-order VSM control algorithm has two configurations.
In [2], the first configuration is presented in detail which is a
voltage-to-current model, and it mainly depends on measuring AC
voltage at PCC. In contrast to [9], a current-to-voltage model was
presented and studied [16], but without energy storage. A
comparison between both high-order control algorithms of VISMA
was studied in [16] under normal operation conditions with
different switching techniques. The control structure of using this
model depends on these equations, so this control strategy can be
done using a hysteresis current control technique or PWM control
technique [16]. Therefore, the voltage-to-current model provides
the reference current values to control the VSC as shown in
Fig. 3a, while the reference voltage values are provided by current-
to-voltage model and can be seen in Fig. 3b.
Where Vg and ig are the voltages at PCC and the output
converter's currents; the excitation voltage of the SM is denoted by
Vf. the symbol wg represents the grid angular speed, and the Pref is
the command reference. The Vref and iref are the output of the SM
which are used as input signals to the converter controllers.
2.2 Low-order VSM model
The high-order VSM control algorithm described above provides
the same properties as SM does. However, the low-order VSM
control algorithm is equivalent to the conventional droop
mechanism; which mainly depends on swing equation [15, 11] as
shown in Fig. 4a. As the main target of introducing VSM is to
mimic the SM behaviour, the emulation of SM's inertia and
damping can be captured using just the swing equation. Therefore,
the general form of the swing equation for SM based on Newton's
law is written in terms of torque parameters as shown in (6) [25],
and also it can be written based on active power in case of
implementing the VSM control by multiplying (6) by the rotor
synchronous speed ωSM as shown in (7) [26]
d2δ dδ
J + Dd = τm − τele (6)
dt2 dt
d2δ dδ
J ⋅ ωVSM ⋅ + DωVSM = Pm − Pele (7)
dt2 dt
According to [1], the derivative term of the rotor angle dδ/dt =
Δω = ωvsm−ωg is the rotor speed deviation in electrical radians per
second; thus it is more convenient to replace the second-order
differential equation term in (7) by two first-order equations as the
following:
dΔω
M = Pmech − Pele − (ωvsm − ωg) Dd (8)
dt
dδ
= Δω (9)
dt
The phase angle and angular speed are used to convert the voltages
and the currents in d–q vector control frame [15].Where the inertia
is denoted by J; the mechanical input power is denoted by Pmech
which is the reference power command, and the electrical power is
represented by Pele. The damping coefficient is denoted by Dd. The
symbol wvsw represents the synchronous speed that is generated by
the swing equation while wg is the estimated grid angular
frequency via phase-locked loop (PLL).
The VSM can be implemented on an inverter with a classical
cascaded voltage and current control loops. Even though, the VSM
based on the swing power equation is simple, the required
behaviour of the inertia and damping is available. The outcomes of
the swing equation are a phase voltage angle θ and angular speed
w.
The cascaded controller loops of the VSC voltage source
converter consist of two stages: the inner and the outer controller
IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769
© The Institution of Engineering and Technology 2017
Table 1 VSC system parameters VSM control algorithms in order to achieve clear guidance for
Quantity Value Unit choosing the appropriate VSM algorithm.
converter rated power 20 kVA
AC voltage (L-L) rms 208 V 3 Simulation results and analysis
AC side resistance 0.15 Ω The system model used in this study consists of a fully detailed
AC side inductance 2 mH switching VSC model, and whose parameters are set out in Table 1.
AC side capacitance 45 µF A detailed VSM-based controller was built in PSCAD/EMTDC.
switching frequency 2 kHz The reason for building a fully detailed system was to study and
system frequency 60 Hz
compare static and dynamic properties during normal and
abnormal operation, such as that which occurs with sudden load
DC voltage 450 V
changes, unbalanced AC voltages, and short circuits. The two
virtual inertias 0.05 and 0.025 kg m2 VSM control approaches proposed in [2, 15, 27] were considered.
The study presented here was concentrated on four important
factors: the amount of active power, the virtual inertia generated,
[26, 27] and Fig. 4b shows the cascaded control structure of VSC the system frequencies, and the value of the virtual damping. The
[28]. This control strategy is used mostly for VSC applications due comparison of the two VSM algorithms was based on their
to the fact that it produces less voltage harmonics [29]. Also, this dynamic properties, the total voltage and the current harmonic
strategy allows independent control of a reactive and an active distortion, sudden load changes, and unbalanced AC voltages.
power [26, 27]. Moreover, the switching control techniques for both high-order and
The main difference between the high- and low-order VSM low-order VSM are based on PWM control.
control algorithm is the controlled structure of the VSC. Even
though the high-order and low-order VSM provides independently 3.1 Case 1: dynamic properties of VSM algorithms during
an active and a reactive power control, each of these algorithms has load changes
a specific control frame. Hence, there is no doubt that both high-
order and low-order VSM algorithms emulate and mimic the SM This case investigates the similarity between the low- and high-
inertia and damping due to the existence of the swing power order based VSM controllers. The results shown in Fig. 5 have
equation in their control structure. Still it is not clear which type of indicated that both VSM models have almost symmetrical and
these is the most efficient choice. For this reason, this paper identical behaviours in terms of dynamic properties. However, it
presents a comprehensive study between high-order and low-order remains unclear as to which type of VSM control is the most

Fig. 3 Block diagram of simplified high-order SM model

(a) Voltage-to-current model, (b) Current-to-voltage model

Fig. 4 Control block implementation

(a) Voltage-to-current model, (b) Current-to-voltage model

IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769 1765
© The Institution of Engineering and Technology 2017
Fig. 5 Dynamic properties of VSM algorithms
(a) Power response for both VSMs with virtual inertia = 0.05 kg m2, (b) Power response for both VSMs with virtual inertia = 0.025 kg m2, (c) Effect of 0.05 kg m2 of virtual inertia
on the system frequency, (d) Effect of 0.025 kg m2 of virtual inertia on the system frequency, (e) Effect of virtual inertia on the drop in system frequency for J = 0.05 kg m2, (f)
Effect of virtual inertia on the drop in system frequency for J = 0.025 kg m2
Fig. 6 Total harmonic distortion (THD) at the PCC
(a) Voltage harmonics distortion (THD)., (b) Current harmonics distortion (THD)
efficient choice. In other words, using VSM, for example, in
distribution system applications might lead to unstable operation in
the case of unbalanced AC voltages; which is commonly occurred
in distribution system. In this case, both VSM algorithms exhibited
similar power and frequency response behaviours for the same
inertia values (J = 0.05 and J = 0.025 kg m2). The damping
coefficients for the VSM control algorithms are unequal because
the low-order VSM is dependent on just the swing equation
expressed while the virtual stator resistances and inductances in the
high-order VSM affect the damping coefficient. Both algorithms
exhibit similar response behaviours for J = 0.05 and J = 0.025 kg
m2 as displayed in Figs. 5a and b, respectively. It is clear that the
power oscillated once command of the active power took place at t
= 2 s due the small damping coefficient. These figures show that
both VSM algorithms mimicked the behaviour of a real
synchronous generator. Figs. 5c and d display the system frequency
response due to sudden load changes. Although virtual inertia is
replicated in the VSC controller through the use of both algorithms,
the implementation of the low-order VSM is preferable due to its
superior design and simplicity of implementation.
The effect of the virtual inertia and virtual damping coefficient
when the system frequency change in case of island AC microgrid
in response to the J = 0.05 kg m2 and J = 0.025 inertia values is
shown in Figs. 5e and f, respectively. At t = 3 s, the system
frequency decreased within an allowable range: 0.3%. The
response of both VSMs based on the virtual inertia was almost
symmetrical. The low-order VSM is also an equivalent to the SM
frequency droop mechanism, so this algorithm mimics real
practical SM characteristics. In fact, no doubt exists that both VSM
algorithms provide similar properties with almost equal levels of
1766
accuracy when the frequency deviations associated with the high-
order algorithm are considered. Therefore, it is clear that increasing
the virtual inertia helps to support the grid with more active power.
3.2 Case 2: total harmonic distortion at the PCC
This case demonstrates the efficiency of the VSM control
algorithms in terms of power quality. The results in case 1 do not
show any differences between the high- and low-order VSM
performances. Therefore, the purpose of this case is to compare the
efficiency of VSM control algorithms based on the total harmonics
distortion. This case shows the advantage of using the low-order
VSM over the high-order VSM control algorithm. Although this
section of the study has not been carried out previously in
literature, it is necessary from a power quality perspective to
improve the system's quality and performance. For this case, the
voltage and current harmonics distortion at the PCC were measured
in order to compare the control algorithms from a power quality
point of view. As clearly shown in Fig. 6a, the low-order control
algorithm generated more voltage harmonics than the high-order
VSM algorithm, but in low-order model the voltage harmonics is
almost equal to 2.8% compared with the high-order model; which
is almost fixed. Based on Fig. 6a, the total harmonics voltage
distortion using the high-order VSM algorithm was fluctuating
around 2%. It is clear that the high-order model produces non-
characteristic harmonics compared with the low-order model.
On the other hand, the total harmonics current distortion using
the low-order VSM algorithm was roughly 0.55%, but in the high-
order VSM algorithm, it was about 0.9% as shown in Fig. 6b. The
high-order VSM algorithm thus spread the harmonics over the
IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769
© The Institution of Engineering and Technology 2017
Fig. 7 THD at the PCC
(a) Voltage THD of low-order VSM model., (b) Current THD of low-order VSM model., (c) Voltage THD of high-order VSM model, (d) Current THD of high-order VSM model

Fig. 8 Short circuit performance

(a) Active power behaviour of both VSMs under a three-phase-to-ground fault, (b) System frequencies for both VSMs under a three-phase-to-ground fault, (c) Active power
behaviour of both VSMs under a two-phase-to-ground fault

entire voltage and current spectrum. In contrast, the low-order 3.3 Case 3: comparison of short circuit performance
VSM has specific harmonics in voltage and current as well, but
they can be easily filtered due to their deterministic pattern. Fig. 7 Both controller algorithms were investigated in case of the three-
shows the individual harmonic orders for both models. phase-to-ground short circuits that take place in the middle of the
The harmonics order and the output voltage can be determined line that connects the PCC to the main grid for durations of 0.1 s,
from (10) and (11), respectively [30]. The total voltage and current or about five cycles. For this case, the dynamic responses of the
harmonics distortions can be calculated using (12) and (13), active power and the frequencies for both controllers are indicated
respectively [31] in Figs. 8a and b, respectively. It can be seen that for a short circuit
applied at t = 8 s, the power of both VSM models decreased almost
horder = 6k ± 1, ∀k = 1, 2, … to zero and that they delivered quite similar levels of dynamic
(10)
performance. However, the low-order model provided more power
than the high-order model. This difference can be controlled by
V LL1 changing the value of the inertia in the low-order model, and it is
V LLh = (11)
h superior with respect to supporting the system frequency and
improving its stability, as shown in Fig. 8a.
∑∞ 2
k = 2 Vk
It is clear that, for both VSM models, the frequencies were
V THD(%) = × 100 (12) increased due to the short circuit and remained within the standard
V1
range. However, for both VSM models, the frequency increases
can be gradual if changes are made to the damping coefficient in
∑k∞= 2 I k2 the swing equation. After the short circuit is removed, the active
I THD(%) = × 100 (13)
I1 power and the frequency for both VSM models were restored, and
they exhibited similar settling times, as shown in Figs. 8a and b.
where horder represents the characteristic and non-characteristic In fact, the three-phase short circuit is not enough to compare
harmonics distortion. The VTHD and ITHD represent the total the VSM models with each other because based on Fig. 3a the
harmonics distortion for voltage and current. The symbol k high-order VSM model mainly depends on the instantaneous
represents an individual harmonic order. measured voltage at the PCC. Therefore, it is necessary to
investigate the performance of both VSM models under either two-
phase-to-ground or single-phase-to-ground short circuit in order to

IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769 1767
© The Institution of Engineering and Technology 2017
Fig. 9 Unbalanced AC voltage performance
(a) Three-phase voltage unbalanced by a 5% distortion, (b) Effect of a 5% voltage unbalance on active power for both VSMs, (c) Effect of a 25% voltage unbalance on active power
for both VSMs
show a fair comparison. For this reason, the two-phase-to-ground is
carried out in this case to reveal the issue of unstable operation that
is correlated with high-order model of VSM. As shown in Fig. 8c,
the two-phase-to-ground short circuit takes place at t = 8 s. Hence,
the power of the high-order VSM model supplied almost 40% of its
reference power, and it has a quite large oscillatory behaviour
compared with the low-order model. On the other hand, the low-
order model is supplying ∼55% of the reference value as shown in
Fig. 8c. It can be clearly seen that during the two-phase-to-ground
short circuit, low-order VSM model shows better performance over
the high-order model under the short circuits study. Indeed, Fig. 8c
shows the undesirable performance of the high-order VSM model
compared with the low-order VSM model during the two-phase-to-
ground short circuit study. Results indicate that, the low-order
VSM model shows better transient performance. The control
structure of low-order VSM, however, offers limited options to
protect the converter from overcurrent issues, which is a very
complex process in the case of high-order VSM model
implementation. Also the control structure of low-order VSM
offers limited options to protect the converter from overcurrent
issues; which is very complex in case of high-order VSM model
implementation.
3.4 Case 4: unbalanced AC voltage
The VSM control algorithms were compared under unbalanced AC
voltages because of the common occurrence of these voltages [32].
The standards permit a small window of deviation in unbalanced
AC voltages. For example, operating an electrical distribution
system with a voltage unbalanced by more than 5% is not endorsed
by ANSI/NEMA [33] nor by IEEE standards. In the case examined
in this study, the phase voltage unbalance rate (PVUR) was
calculated according to (14), which is based on the IEEE definition
[34]. Fig. 9a shows a 5% unbalanced rate in the phase voltage, a
scenario that is still permitted under the IEEE standard. The active
power for both VSMs oscillated due to the effect of the unbalanced
voltages, as is evident in Fig. 9b
82 × V 2abe + V 2bce + V cae
2 validated using a test system consisting of a detailed switching
PVUR % =
V aver
If the enormous potential for the incorporation of
interconnected renewable resources into an electrical system using
power electronics converters is realised, system stability will be
significantly affected because of the lack of inertia [26]. As well,
commonly used single-phase distribution systems will cause
unbalanced voltages when the loads are not equally distributed.
During unbalanced voltage conditions, the active power usually
oscillates around the reference command value, as indicated in
Fig. 9b.
In the case of a sudden increase in the unbalanced rate of the
phase voltage to 25%, the high-order VSM failed to emulate SM
1768
behaviour, and the average value of the active power was also
decreased compared with that of the low-order VSM algorithm, as
shown in Fig. 9c at t = 4 s. The extreme unbalanced voltage by
25% in distribution system could occur during a short circuit
situation [35].
It is obvious that the low-order VSM algorithm is more reliable
than the high-order VSM control algorithm. In addition, according
to [11], using a high-order model to emulate the inertia and control
VSC might lead to numerical problems due to dependence
primarily on measurements of the AC voltages, which is not ideal.
The stability of both control algorithms was examined based on an
unbalanced AC voltage, which is a usual occurrence in electrical
systems. It is obvious that the low-order VSM based on a
frequency droop mechanism control algorithm is the more
promising algorithm and remains stable even during the abnormal
conditions that are common in electrical distribution systems. The
cascaded voltage and current control associated with the low-order
VSM protect the converter from problems arising from
overloading, and it also has a feature that can restrict the power
flow direction.
4 Conclusion
The concept underlying VSM control algorithms has been
reviewed with the goal of providing an efficient and stable VSC
control algorithm that features the emulation of a classical SM.
This paper has categorised a variety of VSM implementations in
terms of additional control loops or equipment required and the
simplicity of the control structures. It has been demonstrated
during normal and abnormal operating conditions. A low-order
VSM algorithm is more stable than a high-order VSM algorithm.
The possibility that the VSM will be unstable during abnormal
operating conditions was therefore studied in order to investigate
the numerical instability that might occur in a practical
implementation. Due to its modified traditional droop control
characteristics, the low-order VSM proved to be a valuable and
useful option that enables a VSC to mimic SM behaviour. The
impact of unbalanced AC voltage on VSM control algorithms was
(14) VSC terminal, simulated in a PSCAD/EMTDC environment.
5. Acknowledgments
The first author, Hasan Al-Rajhi Al-Siraji, acknowledges the
funding support received from Umm AlQura University, Makkah,
Saudi Arabia.
6. References
[1] Bayat, M., Sheshyekani, K., Hamzeh, M., et al.: ‘Coordination of distributed
energy resources and demand response for voltage and frequency support of
MV microgrids’, IEEE Trans. Power Syst., 2016, 31, (2), pp. 1506–1516
IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769
© The Institution of Engineering and Technology 2017
[2] Chen, Y., Hesse, R., Turschner, D., et al.: ‘Improving the grid power quality
using virtual synchronous machines’. IEEE Int. Conf. on Power Engineering,
Energy and Electrical Drives , 2011, pp. 1–6
[3] Slootweg, J.G., Kling, W.L.: ‘Impacts of distributed generation on power
system transient stability’. IEEE Power Engineering Society Summer
Meeting , 2002, pp. 862–867
[4] Bevrani, H., Ise, T., Miura, Y.: ‘Virtual synchronous generators: a survey and
new perspectives’, Int. J. Electr. Power Energy Syst., 2014, 54, (1), pp. 244–
254
[5] Shintai, T., Miura, Y., Ise, T.: ‘Oscillation damping of a distributed generator
using a virtual synchronous generator’, IEEE Trans. Power Deliv., 2014, 29,
(2), pp. 668–676
[6] Van Wesenbeeck, M.P.N., De Haan, S.W.H., Varela, P., et al.: ‘Grid tied
converter with virtual kinetic storage’. IEEE Bucharest PowerTech Conf.,
2009, pp. 1–7
[7] Alipoor, J., Miura, Y., Ise, T.: ‘Power system stabilization using virtual
synchronous generator with alternating moment of inertia’, IEEE J. Emerg.
Sel. Top. Power Electron., 2015, 3, (2), pp. 451–458
[8] Chen, Y., Hesse, R., Turschner, D., et al.: ‘Dynamic properties of the virtual
synchronous machine (VISMA)’. Proc. ICREPQ 2011, pp. 1–5
[9] Zhong, Q.C., Weiss, G.: ‘Synchronverters: inverters that mimic synchronous
generators’, IEEE Trans. Ind. Electron., 2011, 58, (4), pp. 1259–1267
[10] Dong, S., Chi, Y., Li, Y.: ‘Active voltage feedback control for hybrid
multiterminal HVDC system adopting improved synchronverters’, IEEE
Trans. Power Deliv., 2016, 31, (2), pp. 445–455
[11] Arco, S.D., Suul, J.A.: ‘Virtual synchronous machines – classification of
implementations and analysis of equivalence to droop controllers for
microgrids’. PowerTech, 2013, pp. 1–7
[12] Liu, J., Miura, Y., Ise, T.: ‘Comparison of dynamic characteristics between
virtual synchronous generator and droop control in inverter-based distributed
generators’, IEEE Trans. Power Electron., 2016, 31, (5), pp. 3600–3611
[13] Tande, J.O.: ‘On inertial dynamics of virtual-synchronous-controlled DFIG-
based wind turbines’, IEEE Trans. Energy Convers., 2015, 30, (4), pp. 1691–
1702
[14] Wang, S., Hu, J., Yuan, X.: ‘Virtual synchronous control for grid-connected
DFIG-based wind turbines’, IEEE J. Emerg. Sel. Top. Power Electron., 2015,
3, (4), pp. 932–944
[15] Arco, S.D., Suul, J.A.: ‘Equivalence of virtual synchronous machines and
frequency-droops for converter-based microGrids’, IEEE Trans. Smart Grid,
2014, 5, (1), pp. 394–395
[16] Chen, Y., Hesse, R., Turschner, D., et al.: ‘Comparison of methods for
implementing virtual synchronous machine on inverters’. Int. Conf. on
Renewable Energies and Power Quality, 2012, pp. 1–6
[17] Chen, Y., Hesse, R., Turschner, D., et al.: ‘Investigation of the virtual
synchronous machine in the island mode’. IEEE PES Innovative Smart Grid
Technologies Conf. Europe, 2012, pp. 1–6
[18] Hesse, R., Turschner, D., Beck, H.: ‘Micro grid stabilization using the virtual
synchronous machine (VISMA)’. Int. Conf. on Renewable Energies and
Power Quality, 2009, pp. 1–6
[19] Arco, S.D., Are, J., Fosso, O.B.: ‘A virtual synchronous machine
implementation for distributed control of power converters in SmartGrids’,
Electr. Power Syst. Res., 2015, 122, pp. 180–197
[20] Soni, N., Doolla, S.: ‘Inertia design methods for islanded microgrids having
static and rotating energy sources’, IEEE Trans. Ind. Appl., 2016, PP, (99),
pp. 1–10
[21] Yu, M., Roscoe, A.J., Booth, C.D., et al.: ‘Use of an inertia-less virtual
synchronous machine within future power networks with high penetrations of
converters’. Power Systems Computation Conf. (PSCC), 2016, pp. 1–7
[22] Suul, J.A., Arco, S.D., Guidi, G.: ‘Virtual synchronous machine-based control
of a single-phase bi-directional battery charger for providing vehicle-to-grid
services’, IEEE Trans. Ind. Appl., 2016, 52, (4), pp. 3234–3244
[23] Guan, M., Pan, W., Zhang, J., et al.: ‘Synchronous generator emulation
control strategy for voltage source converter (VSC) stations’, Power Syst.
IEEE Trans., 2015, pp, (99), pp. 1–9
IET Gener. Transm. Distrib., 2017, Vol. 11 Iss. 7, pp. 1762-1769
© The Institution of Engineering and Technology 2017
View publication stats
[24] Etemadi, A.H., Iravani, R.: ‘Supplementary mechanisms for smooth transition
between control modes in a microgrid’, Electr. Power Syst. Res., 2017, 142,
pp. 249–257
[25] Machowski, J., Bialek, J.W., Bumby, J.R.: ‘Power system dynamics stability
and control’ (Wiley, 2008, 2nd edn.)
[26] D'Arco, S., Suul, J.A., Fosso, O.B.: ‘Small-signal modeling and parametric
sensitivity of a virtual synchronous machine in islanded operation’, Electr.
Power Energy Syst., 2015, 72, pp. 3–15
[27] Arco, S.D., Suul, J.A., Fosso, O.B.: ‘Automatic tuning of cascaded controllers
for power converters using eigenvalue parametric sensitivities’, IEEE Trans.
Ind. Appl., 2015, 51, (2), pp. 1743–1753
[28] Alsiraji, H.K.A., El-saadany, E.F.: ‘Cooperative autonomous control for
active power sharing in multi-terminal VSC-HVDC’, Int. J. Process. Syst.
Eng., 2014, 2, (4), pp. 303–319
[29] Alsiraji, H.A.: ‘Cooperative power sharing control in multi- terminal VSC-
HVDC’ (University of Waterloo, 2014)
[30] Mohan, N., Undeland, T., Robbins, W.: ‘Power electronics: converters,
applications, and design’ (John Wiley & Sons, Inc., 2013, 3rd edn.)
[31] Hoevenaars, T., Eng, P., Ledoux, K., et al.: ‘Interpreting IEEE Std 519 and
meeting its harmonic limits in VFD applications’. Proc. of the IEEE Industry
Applications Society 50th Annual Petroleum and Chemical Industry Conf.,
2003, pp. 1–6
[32] Lee, K., Venkataramanan, G., Jahns, T.M.: ‘Source current harmonic analysis
of adjustable speed drives under input voltage unbalance and sag conditions’,
IEEE Trans. Power Deliv., 2006, 21, (2), pp. 567–576
[33] Das, J.C.: ‘Power system analysis: short-circuit load flow and harmonics’
(CRC Press, 2002)
[34] Pillay, P., Manyage, M.: ‘Definitions of voltage unbalance’, IEEE Power Eng.
Rev., 2002, 22, (11), pp. 49–50
[35] Hannan, M.a., Mohamed, A., Member, S., et al.: ‘Dynamic phasor modeling
and EMT simulation of USSC’. Proc. of the World Congress on Engineering
and Computer Science, 2009, pp. 1–7
[36] Pulendran, S., Tate, J.E.: ‘Hysteresis control of voltage source converters for
synchronous machine emulation’. 15th Int. Power Electronics and Motion
Control Conf., 2012, pp. 1–8
[37] Hirase, Y., Abe, K., Sugimoto, K., et al.: ‘A grid-connected inverter with
virtual synchronous generator model of algebraic type’, Electr. Eng. Jpn.,
2013, 184, (4), pp. 10–21
[38] Hirase, Y., Noro, O., Yoshimura, E., et al.: ‘Virtual synchronous generator
control with double decoupled synchronous reference frame for single-phase
inverter’. The Int. Power Electronics Conf., 2014, pp. 1552–1559
[39] Pogaku, N., Prodanović, M., Green, T.C.: ‘analysis and testing of autonomous
operation of an inverter-based microgrid’, IEEE Trans. Power Electron.,
2007, 22, (2), pp. 613–625
[40] Guerrero, J.M., Matas, J., De Vicuña, L.G., et al.: ‘Wireless-control strategy
for parallel operation of distributed-generation inverters’, IEEE Trans. Ind.
Electron., 2006, 53, (5), pp. 1461–1470
[41] Yao, W., Chen, M., Matas, J., et al.: ‘Design and analysis of the droop control
method for parallel inverters considering the impact of the complex
impedance on the power sharing’, IEEE Trans. Ind. Electron., 2011, 58, (2),
pp. 576–588
[42] Vasquez, J.C., Guerrero, J.M., Luna, A., et al.: ‘Adaptive droop control
applied to voltage-source inverters operating in grid-connected and islanded
modes’, IEEE Trans. Ind. Electron., 2009, 56, (10), pp. 4088–4096
[43] Delille, G., François, B., Malarange, G.: ‘Dynamic frequency control support
by energy storage to reduce the impact of wind and solar generation on
isolated power system's inertia’, IEEE Trans. Sustain. Energy, 2012, 3, (4),
pp. 931–939
[44] Beck, H.P., Hesse, R.: ‘Virtual synchronous machine’. 9th Int. Conf. on
Electrical Power Quality and Utilisation, EPQU, 2007, pp. 1–6
1769